**1. Introduction**

The global logistics flows have increased dramatically in recent years due to a globalized economy that introduces inherent challenges to the establishment of international business [1] (p. 10). This evolution is combined with the demands of customers who want to be served with shorter delivery times [2] (p. 1) as well as with the increasing product variants in manufacturing and assembly processes that expose planning and control logistics to new challenges [3] (p. 797). The reasons are mainly the variability of final customers favored by competition activities. Other factors that generate uncertainty effects in the planning within the supply chain are, for example, minimum quantities in production as well as deviations of delivery lead times that make more difficult the production planning reducing the planning quality. The conventional answer to this challenge is to increase safety stocks to ensure the expected service level [4] (p. 179).

The success of the actors involved in a cooperative supply chain depends in an essential way on the extent to which they are able to deal with the dynamic market requirements [1] (p. 101)—together with efficient supply processes in costs, a reliable planning within the supply chain as well as robust production and logistics systems play a fundamental role for the long-term success of the actors within the supply chain. Robust production systems avoid negative effects on production processes by identifying potential disruptions and enabling an early adaptation [5] (p. 575). The result of a survey within a study of the German Association of Logistics ratifies this statement detailing that the reliability of logistics and production systems is the factor with greatest relevance considering logistics costs, reaction capabilities, flexibility and use of resources [6]. The logistics goal of robustness defined as the capacity of a system to deal with breakdowns, deviations and changes of the system environment without necessary changes in the structure or in production capacities [7] (p. 34) is gaining more and more importance as strategic feature to secure competitiveness.

Information in real time and its processing for production planning and control can give the correct answer how to face increasing dynamic requirements. However, the current planning and control logics do not provide the necessary support for this, which causes unexpected deviations in planning which cannot be compensated in the short term [7] (p. 160). The turbulences in production and in the supply chain create uncertainties that generate the need for coordination of planning processes, which has not been considered in a methodical way for the current concepts of production planning and control [8] (pp. 36–37). In this way, the insufficient collection and use of planning information lead to delays and deficits in the transmission of information that must be compensated through additional costs [9] (pp. 3–5). Moreover, due to the increasing automation and implementation of process control systems, more and more decentralized maintenance units are located next to production areas. Based on these facts, the common task for both production and maintenance must be developed to optimize the availability and condition of production plants to guarantee the satisfaction of the final customer. From a production point of view, maintenance has evolved from the "auxiliary need" to the holistic maintenance management, in which production and maintenance share a common goal [10] (p. 9).

According to a survey carried out by the VDMA (Mechanical Engineering Industry Association) organization of 240 European companies of all sizes and of different industries, in 2006, 78.5% of managers claimed that the importance of maintenance has increased significantly in recent years and 67.1% that it will continue to increase in the future [11] (p. 38). Considering this, it is prioritized between maintenance and production tasks. The negative long-term effects of insufficiently implemented maintenance measures are often accepted due to short-term production needs. Without denying the character of service, the role of maintenance is different if it is more widely understood as holistic maintenance management. Maintenance management is an essential component of production management but should not be used only to meet the production objectives (production of goods and services), but also the objectives of generating value, human objectives and environmental aspects of the company [10] (pp. 9–10).

Many approaches have been considered to make production and maintenance planning as efficient and flexible as possible. All pursue and many theoretically achieve the partial or total optimization of global production systems by making them efficient. However, when applying these theoretical internal logistics management models to practice, information systems should be used as a vehicle to communicate the necessary information to be managed within the organization. The reasons why the majority of applications of such concepts in information systems, as in ERP (Enterprise Resource Planning) systems, have failed are: the delay or lack of information, the misuse of production planning tools and maintenance, lack of coordination or conflict of interest between the production and maintenance departments, the non-consideration of the environment and the requirements of the final customer in a dynamic process of continuous improvement, leading all of them to take strategic, tactical or operational decisions at an optimal time or in an optimal way.

Research focus is the design and simulation of a coordination model for production and maintenance management using as a basis the current state of the art and pursuing high adaptability to environment changes. To achieve this, the Viable System Model (VSM) is used as a methodological structured approach and system dynamics is used to describe interrelationships between parameters and control loops in production and maintenance. The goal of the research is to generate an approach of how to design and develop production systems to be successful in all potential future environment scenarios considering developments in other areas of the company and in external fields of a company such as customer and supplier needs and conditions, politics, economy, technology, environment and energy regulations and society. In this context, simulation serves as tool to prove the approach in some potential future scenarios for production and maintenance coordination.

The initial hypothesis is that a production system built using the structure of the VSM will be able to react faster to environment changes and therefore to improve short-, medium- and long-term goals of every producing company.

#### **2. Materials and Methods**

This chapter contains a literature review of the viable system model, system dynamics as well as a description of the simulation technique and software employed.

#### *2.1. The Viable System Model (VSM)*

The Viable System Model (VSM), a cybernetic management model, was developed by Stafford Beer throughout his life [12] (p. 57). Beer deduced the VSM by taking the central nervous system of the human being and cybernetics as a basis to deal with complex systems. The VSM is built on three main principles: viability, recursivity and autonomy [13] (p. 434). The cybernetic model of every viable system consists always in a structure with five necessary and sufficient subsystems that are in relation in any organism or organization with the ability to conserve its identity with independency of its environment [12] (pp. 21–22).

To validate the research methodology, research and practical applications using the VSM were searched. Many authors have used the VSM as basis to describe and develop models how to deal with complex challenges in social and industry fields. Some of the topics worked and that give an indication of the scientific value of the approach are: organizational models for companies [14] (pp. 74–76), lean methods in terms of attenuating and amplifying variety [15], production system focused on "make-to-order" manufacturing [16], optimization of patient care in university hospitals [17], order booking process in mass production companies [18], production master program during launch processes [19], production planning in real time in the consumer goods industry [20] and integrated planning of distribution networks [21].

As described in the literature, the VSM is an unmatched conceptual and methodological tool for the modeling and design of organizations and its areas with the goal of being viable [22] (p. 16). Thus, the aim of the research is to propose a self-regulating approach of how designing and coordinating production and maintenance within manufacturing companies. For this reason, the Viable System Model is applied for this purpose.

### *2.2. System Dynamics (SD)*

System dynamics is a computer-guided approach for studying, managing and solving complex feedback problems with focus on policy analysis and design [23] (p. 342). The origin of system dynamics [24] is the field developed by Forrester called "Industrial Dynamics" [25]. It proposes a methodology for the simulation of dynamic models by studying the characteristics of the information feedback of industrial systems.

SD has been applied to a great set of systems from corporate strategy to the dynamics of diabetes as well as for the cold war arms race between the USA and USSR. System dynamics can be applied to any dynamic system, with any time and spatial scale [24] (pp. 41–42). In a firm context, SD addresses three important issues: it helps to determine which policies should be used to control the behavior of the firm over time and, when the circumstances change, how these policies should be designed in order to have a robust response against change and how the organization can design its information feedback structure to assure the correct implementation of effective policies [26] (p. 3). The use of System Dynamics Modeling in Supply Chain Management has re-emerged in recent years after a long-stagnated period [23] (p. 342). The application of systems dynamics for the coordination of production management and maintenance makes sense since the cause–effect relationships show the interrelationships between the elements of the system and help to evaluate the influences of the different decisions in the global system. Therefore, it is selected for the research purpose.

### *2.3. Simulation Software*

According to "VDI-Richtlinie", simulation is "the reproduction of a system with its dynamic processes in an experimental model capable to gain knowledge that can be transferred to reality" [27] (p. 48). Simulation models are mainly used to support decision-making because they show the dynamic behavior of a system [28] (p. 28). Simulation is the only practical way to test models because our mental models are dynamically deficient, omitting feedback, time delays, accumulations and nonlinearities [24] (p. 37). For all these reasons, simulation is used to reproduce the conceptual model and to validate initial hypotheses. In the market, there are different software packages that enable system dynamics modeling such as: AnyLogic, DYNAMO, iTHINK, POWERSIM, STELLA and VENSIM [29] (p. 43). From all of them, VENSIM simulation software was selected for the research work. VENSIM is a registered trademark of Ventana Systems Inc., Harvard, MA, USA) serves as platform to build stock and flow model diagrams as well as causal loop diagrams. VENSIM also provides very powerful tools for analysis and validation of results and model structure as well as to determine most convenient policy options.

#### **3. Literature Review on Production and Maintenance Management**

This chapter provides the basic terminology needed to successfully perform this work.

#### *3.1. Production Management*

Production is the foundation of human activity. Natural resources are transformed into useful products through production processes to meet the needs of society [30] (p. 319). Production management contains the tasks of design, planning, monitoring and control of the productive system and business resources such as people, machines, material and information [31] (pp. 249–273). The multi-dilemma of production planning originates discussions repeatedly in the context of divergent objectives. This conflict of goals is shown in Figure 1 [32] (p. 36).

**Figure 1.** Multi-dilemma of production planning [32] (p. 36).

To analyze the tasks of production management, the Aachener Production Planning and Control (PPC) model, which is a reference model for its analysis, evaluation and design, is used [33] (p. 29).

All tasks are distinguished vertically in Figure 2 according to their strategic, tactical or operational nature [33] (pp. 30–32). In the past, the focus was on operational and tactical problems; however, to successfully manage logistics in the future, an active level of strategic planning is also required [34] (p. 1).


**Figure 2.** Production management tasks acc. to the Aachener PPC [33] (p. 30).

#### *3.2. Maintenance Management*

Industrial maintenance is defined according to DIN (German Institute for Standardization) 31051 as the "combination of all technical, administrative and management measures during the life cycle of an observation unit in order to maintain the functional status or restoring it so that it can fulfill the required function" [35] (p. 26). The greater degree of complexity of technologies, systems and processes has considerably increased the demands on operating time. The increase of maintenance importance is a logical consequence since maintenance processes affect the quality, delivery time and costs and therefore company's performance [15] (pp. 16–17). An optimal strategy must be found between the costs of preventive maintenance and the costs of machine failures [35] (p. 25). The basic maintenance strategies are corrective, periodic, preventive oriented to the condition of the object and predictive [35] (p. 104). The three "classic" maintenance strategies are still justified as seen in Figure 3.

**Figure 3.** Survey result of 240 European companies on maintenance strategies [11] (p. 41).

The "ideal maintenance organization" does not exist [35] (p. 65) but depends on optimally combining the organizational forms with their advantages and disadvantages for the respective company [11] (p. 23).

According to experts, German companies spend around 140,000 million euros annually on maintenance of machines and installations. This, together with the fact that between 75% and 80% of the execution time of a maintenance order are activities without added value, defines a great potential for savings in production companies. Many of the time losses are in the interfaces between the different areas. The management of maintenance processes must ensure the change from a functional orientation to an orientation towards processes [35] (p. 60). For this purpose, tasks of maintenance management are: planning, control, analysis and measures execution [36] (p. 173).

## *3.3. Coordination Concepts of Production and Maintenance Management*

Maintenance management of production units affects not only the maintenance personnel but also all the employees of the whole company [35] (p. 104). In the last two decades, there has been a general rethinking of industrial maintenance philosophy, from the maintenance of functions to a philosophy of value creation. However, for many maintenance tasks, there is a lack of adequate methods and instruments as well as information technology solutions [11] (p. 69).

Logistics can be described as an interdisciplinary strategy to optimize production. Availability is, therefore, a key factor for logistics and is also the link between logistics and industrial maintenance [35] (pp. 4–6). Maintenance should be "integrated" and dependent on all the functional areas involved in the added value [36] (pp. 9–10). This maintenance concept refers to the function and not just to the maintenance department within an organization. This is characterized by a distribution of the maintenance function "on several shoulders" [11] (p. 38). For all the above, some of the approaches to integrate or manage maintenance within the productive system are [36] (pp. 4–9):


### **4. Results**

Within the research work, a conceptual model is developed and simulated for the integrated planning of production and maintenance based on the Viable System Model (VSM).

#### *4.1. Conceptual Model Design Applying the VSM*

#### 4.1.1. Methodology

This chapter describes the components of an integrated planning model for production and maintenance management. In a first step, the planning tasks are presented according to its planning level (strategic, tactical, and operational). Later, the recursion levels of the VSM are described. Then, the planning tasks are associated with the different recursion levels and the information flows are defined between recursion levels and systems of the VSM.

#### 4.1.2. Production and Maintenance Management Tasks Acc. to Their Planning Horizons

Production systems are considered important in relation to aspects of quality, time and costs [37] (p. 1). As explained before, planning tasks can be classified into strategic, tactical and operational planning depending on the respective planning horizon. Therefore, this classification was performed for the production management tasks in Figure 4. Moreover, with the increase of automation levels and the decrease of personnel in production, the importance of maintenance is increasing. With the same methodology, maintenance management tasks are also classified into the different time horizons in Figure 5:


**Figure 4.** Production management and planning tasks acc. to time horizons (own elaboration).


**Figure 5.** Maintenance management and planning tasks acc. to time horizons (own elaboration).
